Wiki-CS: A Wikipedia-Based Benchmark for Graph Neural Networks

Abstract

We present Wiki-CS, a novel dataset derived from Wikipedia for benchmarking Graph Neural Networks. The dataset consists of nodes corresponding to Computer Science articles, with edges based on hyperlinks and 10 classes representing different branches of the field. We use the dataset to evaluate semi-supervised node classification and single-relation link prediction models. Our experiments show that these methods perform well on a new domain, with structural properties different from earlier benchmarks. The dataset is publicly available, along with the implementation of the data pipeline and the benchmark experiments, at https://github.com/pmernyei/wiki-cs-dataset .

arXiv commentGraph Representation Learning and Beyond workshop (ICML 2020); corrected incorrect metrics due to issue in experiment implementation

References in corpus (6)

• Variational Graph Auto-Encoders
• Pitfalls of Graph Neural Network Evaluation
• Open Graph Benchmark: Datasets for Machine Learning on Graphs
• GraphSAINT: Graph Sampling Based Inductive Learning Method
• Benchmarking Graph Neural Networks
• Graph Star Net for Generalized Multi-Task Learning

Cited by in corpus (19)

• Graph Contrastive Learning with Adaptive Augmentation
• Graph Self-Supervised Learning: A Survey
• Graph Barlow Twins: A self-supervised representation learning framework for graphs
• How to Find Your Friendly Neighborhood: Graph Attention Design with Self-Supervision
• Graph Communal Contrastive Learning
• MagNet: A Neural Network for Directed Graphs
• Understanding and Resolving Performance Degradation in Graph Convolutional Networks
• Automated Self-Supervised Learning for Graphs
• GraphMixup: Improving Class-Imbalanced Node Classification on Graphs by Self-supervised Context Prediction
• Geometric Scattering Attention Networks
• Benchmarking Graph Neural Networks on Link Prediction
• An Empirical Study of Graph Contrastive Learning
• Evaluating Modules in Graph Contrastive Learning
• Embedding Graphs on Grassmann Manifold
• Neuron with Steady Response Leads to Better Generalization
• Graph Denoising with Framelet Regularizer
• Self-Supervised Graph Learning with Proximity-based Views and Channel Contrast
• Curvature Graph Neural Network
• Interferometric Graph Transform for Community Labeling